I should begin by saying that I like this book, that I enjoy Mr. Petroski's writing style, and agree with his premise -- that we stand to learn more by studying a single failure than a thousand successes. Mr. Petroski makes an ample case for this through the judicious use of historic failures, some of which are more historic than others (one example, from ancient Greece, involves methods of storing marble columns). To his credit, Mr. Petroski's writing style is approachable by non-engineers, a feat that is probably worth at least one star all by itself. But it is a shortcoming that considerable detail has been sacrificed, perhaps in an effort to make the text palatable to a non-technical audience. The resulting text glosses over mechanical reasons for the design flaws under consideration. In some instances, such details are probably not all that important. To be fair, lengthy technical explanations about collapsed bridges, broken ships or fractured colums might render this book even less marketable than it is (at present, it hovers below 14,000th on Amazon's sales ranking). In those cases, the omission simply makes the account less satisifying to the overly curious reader. But that is not always the case, and some examples would have benefitted from more detailed explanations for two reasons. First, since the book is about learning from mistakes, it would have been valuable to understand the mistake itself. That knowledge would help the reader appreciate how subsequent engineers evaluated a problem, identifed its cause and avoided repeating the mistake in analogous situations. Second, and more troubling, some omissions are confusing. For instance, the Challenger disaster is compared to the aforementioned Greek column problem. In the former, a third O-ring had been added to the shuttle engine, in the latter a third brace used to hold up the columns, and Mr. Petroski offers this as a "red flag" that should have alerted NASA engineers to the need to reevaluate the whole system. I'm no engineer, but my understanding of what caused the failure of the shuttle's third O-ring is different in kind from the problem associated with third brace used in the storage of marble columns. On balance, however, these are not fatal flaws. Mr. Petroski's book is worth reading by anyone who rides an elevator, works in a skyscraper or drives across a bridge. And his central point -- that system modifications justify a reevaluation of the entire system for unintended design problems -- is one that should be taken to heart by engineers and non-engineers alike.